def tag_sentences(self, text, hyp, verbose=False):
"""
Tag a RTEPair as to whether the hypothesis can be inferred from the text.
"""
text_drs_list = drt_glue.parse_to_meaning(text, dependency=True)
if text_drs_list:
text_ex = text_drs_list[0].simplify().toFol()
else:
if verbose: print 'ERROR: No readings were be generated for the Text'
hyp_drs_list = drt_glue.parse_to_meaning(hyp, dependency=True)
if hyp_drs_list:
hyp_ex = hyp_drs_list[0].simplify().toFol()
else:
if verbose: print 'ERROR: No readings were be generated for the Hypothesis'
#1. proof T -> H
#2. proof (BK & T) -> H
#3. proof :(BK & T)
#4. proof :(BK & T & H)
#5. satisfy BK & T
#6. satisfy BK & T & H
result = inference.get_prover(hyp_ex, [text_ex]).prove()
if verbose: print 'prove: T -> H: %s' % result
if not result:
bk = tagger._generate_BK(text, hyp, verbose)
bk_exs = [bk_pair[0] for bk_pair in bk]
if verbose:
print 'Generated Background Knowledge:'
for bk_ex in bk_exs:
print bk_ex.infixify()
result = inference.get_prover(hyp_ex, [text_ex]+bk_exs).prove()
if verbose: print 'prove: (T & BK) -> H: %s' % result
if not result:
consistent = self.check_consistency(bk_exs+[text_ex])
if verbose: print 'consistency check: (BK & T): %s' % consistent
if consistent:
consistent = self.check_consistency(bk_exs+[text_ex, hyp_ex])
if verbose: print 'consistency check: (BK & T & H): %s' % consistent
return result
def demo_drt_glue_remove_duplicates(show_example=-1):
from nltk_contrib.gluesemantics import drt_glue
examples = ['David sees Mary',
'David eats a sandwich',
'every man chases a dog',
'John chases himself',
'John likes a cat',
'John likes every cat',
'he likes a dog',
'a dog walks and he leaves']
example_num = 0
hit = False
for sentence in examples:
if example_num==show_example or show_example==-1:
print '[[[Example %s]]] %s' % (example_num, sentence)
readings = drt_glue.parse_to_meaning(sentence, True)
for j in range(len(readings)):
reading = readings[j].simplify().resolve_anaphora()
print reading
print ''
hit = True
example_num += 1
if not hit:
print 'example not found'
def _exampleList_select(self, event):
selection = self._exampleList.curselection()
if len(selection) != 1: return
self._curExample = int(selection[0])
example = self._examples[self._curExample]
if example:
self._exampleList.selection_clear(0, 'end')
self._exampleList.selection_set(self._curExample)
self._readings = drt_glue.parse_to_meaning(example, self._remove_duplicates)
self._populate_readingListbox()
self._drs = None
self._redraw()
else:
# Reset the example selections.
self._exampleList.selection_clear(0, 'end')
def _exampleList_select(self, event):
selection = self._exampleList.curselection()
if len(selection) != 1: return
self._curExample = int(selection[0])
example = self._examples[self._curExample]
if example:
self._exampleList.selection_clear(0, 'end')
self._exampleList.selection_set(self._curExample)
self._readings = drt_glue.parse_to_meaning(example,
self._remove_duplicates)
self._populate_readingListbox()
self._drs = None
self._redraw()
else:
# Reset the example selections.
self._exampleList.selection_clear(0, 'end')
def demo_inference_tagger(verbose=False):
tagger = RTEInferenceTagger()
text = 'John see a car'
print 'Text: ', text
hyp = 'John watch an auto'
print 'Hyp: ', hyp
# text_ex = logic.LogicParser().parse('some e x y.((david x) and ((own e) and ((subj e x) and ((obj e y) and (car y)))))))')
# hyp_ex = logic.LogicParser().parse('some e x y.((david x) and ((have e) and ((subj e x) and ((obj e y) and (auto y)))))))')
text_drs_list = drt_glue.parse_to_meaning(text, dependency=True, verbose=verbose)
if text_drs_list:
text_ex = text_drs_list[0].simplify().toFol()
else:
print 'ERROR: No readings were be generated for the Text'
hyp_drs_list = drt_glue.parse_to_meaning(hyp, dependency=True, verbose=verbose)
if hyp_drs_list:
hyp_ex = hyp_drs_list[0].simplify().toFol()
else:
print 'ERROR: No readings were be generated for the Hypothesis'
print 'Text: ', text_ex
print 'Hyp: ', hyp_ex
print ''
#1. proof T -> H
#2. proof (BK & T) -> H
#3. proof :(BK & T)
#4. proof :(BK & T & H)
#5. satisfy BK & T
#6. satisfy BK & T & H
result = inference.get_prover(hyp_ex, [text_ex]).prove()
print 'prove: T -> H: %s' % result
if result:
print 'Logical entailment\n'
else:
print 'No logical entailment\n'
bk = tagger._generate_BK(text, hyp, verbose)
bk_exs = [bk_pair[0] for bk_pair in bk]
print 'Generated Background Knowledge:'
for bk_ex in bk_exs:
print bk_ex.infixify()
print ''
result = inference.get_prover(hyp_ex, [text_ex]+bk_exs).prove()
print 'prove: (T & BK) -> H: %s' % result
if result:
print 'Logical entailment\n'
else:
print 'No logical entailment\n'
# Check if the background knowledge axioms are inconsistant
result = inference.get_prover(assumptions=bk_exs+[text_ex]).prove()
print 'prove: (BK & T): %s' % result
if result:
print 'Inconsistency -> Entailment unknown\n'
else:
print 'No inconsistency\n'
result = inference.get_prover(assumptions=bk_exs+[text_ex, hyp_ex]).prove()
print 'prove: (BK & T & H): %s' % result
if result:
print 'Inconsistency -> Entailment unknown\n'
else:
print 'No inconsistency\n'
result = inference.get_model_builder(assumptions=bk_exs+[text_ex]).build_model()
print 'satisfy: (BK & T): %s' % result
if result:
print 'No inconsistency\n'
else:
print 'Inconsistency -> Entailment unknown\n'
result = inference.get_model_builder(assumptions=bk_exs+[text_ex, hyp_ex]).build_model()
print 'satisfy: (BK & T & H): %s' % result
if result:
print 'No inconsistency\n'
else:
print 'Inconsistency -> Entailment unknown\n'
